Protecting the innocent in eyewitness identification: An analysis of simultaneous and ranking lineups

Abstract

In an effort to protect innocent suspects in police lineups, guidelines tend to encourage conservative responding in eyewitnesses. We used Signal Detection Theory (SDT), in conjunction with Expected Information Gain (EIG), to explain why conservative responding with standard simultaneous lineup procedures is detrimental to gathering information about the guilt or innocence of suspects. We also show that a different lineup procedure, the ranking lineup, should largely avoid this loss of information. These SDT predictions were tested in two experiments that manipulated response conservativeness in terms of instructions to the witness and/or witness confidence levels. The results showed strong evidence for the predicted pattern. That is, conservative responding substantially decreased the information value of witness responses in simultaneous lineups, but not ranking lineups. Critically, conservative responding in the simultaneous procedure specifically decreased the ability to gain evidence of innocence, revealing a cost that offsets the benefit of reduced false identifications. The ranking procedure, in contrast, provided strong evidence of innocence even when false identification rates were low. These results have significant implications for policy recommendations in police lineups and suggest that eyewitness researchers need to consider information-theory measures in the attempt to find procedures that best serve the goal of protecting innocent suspects.